Prevailing practices in evaluating workfare programs have ignored the disutility of the type of work done, with theoretically ambiguous implications for the impacts on poverty. In the case of India's National Rural Employment Guarantee Scheme, past assessments have relied solely on household consumption per person as the measure of economic welfare. The paper generalizes this measure to allow for the disutility of casual manual work. The new measure is calibrated to the distribution of the preference parameters implied by maximization of an idiosyncratic welfare function assuming that there is no rationing of the available work. The adjustment implies a substantially more "poor-poor" incidence of participation in the scheme than suggested by past methods. However, the overall impacts on poverty are lower, although still positive. The main conclusions are robust to a wide range of alternative parameter values and to allowing for involuntary unemployment using a sample of (self-declared) un-rationed workers.